R. W. H. ROW—REPORT ON THE SPONGES: CALCAREA. 193 
The series Leucandra—Leucilla is an eminently natural one, species like 
Leucandra verdensis, Thacker (18) and Leucilla intermedia, n. sp., described 
below, being directly intermediate between the typical Leucandras, which 
have no subdermal quadriradiates,and the typical Leucillas, in which they are 
extremely large. L. verdensis, although possessing subdermal quadriradiates, 
is placed in the genus Leucandra by Thacker on account of the “ inconspi- 
cuousness ” of these spicules (cf. diagnosis of the family Amphoriscide in 
Dendy (2)), in comparison with the dermal tangentially lying triradiates ; 
while my new species L. intermedia, in which the quadriradiates are but little 
larger, has been placed in the genus Lewcilla since the dermal triradiates are 
by no means so large as the quadriradiates, which are therefore “conspicuous.” 
In both these species the apical rays are small in comparison with the facial 
rays. On the other hand, it is obviously impossible to derive such genera 
as Syculmis and Amphoriscus, in which the canal-system is syconoid or 
sylleibid, from any form which has, like Leweandra, a typically leuconoid 
canal-system. It is therefore suggested that the “ Amphoriscidz Sycones ” 
are derived from the Grantillide, and the “ Amphoriscidee Leucones” from 
the leuconoid Grantidee. 
The Heteropide, which like the Amphoriscidee are descended from primitive 
Grantillid ancestors, possess features far more difficult of explanation than do 
the Amphoriscide. Chief amongst these is the occurrence of subdermal 
sagittal triradiates and prochiacts. 
It is impossible to derive these spicules from any tangential dermal spicules 
merely by the addition of a ray, as in the case of the quadriradiates, and they 
must therefore have been developed either as a completely new structure, or 
by.a change of position of one or more of the rays of an existing spicule. 
On behalf of the second of these two views, that these spicules are developed 
from dermal triradiates by a bending of the basal ray, there is strong pre- 
sumptive evidence to be obtained from the study of the development of the 
chiactines of the Chiphoridee and the Staurorrhaphide. 
Tn six species out of nine Jenkin states that the chiactines, which do not 
extend right up to the osculum, are replaced in that region by ordinary gastral 
quadriradiates, whose facial rays lie in the primitive facial plane and whose 
apical rays point into the gastral cavity. Between the areas in which these 
two types occur, there is an intermediate region in which quadriradiates are 
found “in all intermediate positions between tangential and centrifugal,” 7. e., 
with the basal ray directed in all intermediate positions between aborally, as 
in the gastral quadriradiate, and dermalwards, as in the typical chiact. This 
change of position, he suggests, is caused by the appearance of the flagellated 
chambers, which continually appear in that region as the sponge elongates, so 
that “the spicules formed in the oscular collar may be supposed to be turned 
round by the development of flagellated chambers under their basal rays. A 
very similar tipping of dermal triradiates, due to the growth under them of 
